Module 1 Wrap-up: Your First Agentic Decision
Hands-on: Compare a chatbot vs an agent workflow and identify agent-worthy problems.
Module 1 Wrap-up: Designing the Loop
You have learned the theory of agentic AI. Now, we are going to practice the Engineering Mindset. Before you write code, you must be able to design the "Mental Loop" of the agent.
Hands-on Exercise 1: The Chatbot vs. The Agent
The Scenario: You want to know if you should bring an umbrella to your meeting tomorrow at 2 PM in Seattle.
The Chatbot Workflow (Reactive)
- User: "Will it rain in Seattle tomorrow at 2 PM?"
- Model: "I don't have real-time access, but Seattle is often rainy..."
- User: Frustrated, goes to Google.
The Agentic Workflow (Proactive)
- User: "Will it rain in Seattle tomorrow at 2 PM?"
- Agent Thought: "I need real-time weather. I have a
get_weathertool." - Agent Action:
get_weather(city="Seattle", date="2025-12-20") - Observation: "80% chance of rain at 14:00."
- Agent Final Answer: "Yes, bring an umbrella. There is a high chance of rain exactly during your 2 PM meeting."
Hands-on Exercise 2: Identify Agent-Worthy Problems
Look at these three tasks. Which one is for a Script, a Chatbot, or an Agent?
- Task A: Calculating the total of an invoice from 10 line items.
- Task B: Writing a 500-word blog post about "The Future of Space Travel."
- Task C: Monitoring a competitor's website and sending a Slack alert if they change their pricing.
The Answer:
- Task A: Script. (Math is deterministic).
- Task B: Chatbot. (One-shot generative task).
- Task C: Agent. (Must explore a dynamic website, reason about what 'prices' are, and take an action on Slack).
Module 1 Summary
- Agentic AI is the transition from "Talking" to "Doing."
- Autonomy is the ability to handle multi-step loops without user input.
- The ReAct Pattern (Thought-Action-Observation) is the foundation of all agents.
- Engineering discipline means choosing the right tool: don't use an agent for a simple function.
Coming Up Next...
In Module 2, we move from theory to components. We will look at exactly what goes inside the "Brain" of an agent, including Memory, Tools, and the critical Control Loop.
Module 1 Checklist
- I can explain the difference between Wave 2 (Generative) and Wave 3 (Agentic) AI.
- I understand why "Temperature 0" is used in agents.
- I can draw a basic ReAct loop.
- I identified a problem in my own work that is "Agent-worthy."
- I understand the latency and cost trade-offs of agents.